Well, 4f is an electron from a 4f electron, but I guess you knew that part, already, right?
The indices 7/2 and 5/2 indicate that the electrons exhibit coupled states thus resulting in a split signal. For any angular momentum or spin phenomenon multiplicities of coupled states are calculated by 2J+1, that means you should observe two 4f signals whose intensities should have the ratios:
7/2*2+1=8
5/2*2+1=6
->8/6=4/3=I(4f7/2)/I(4f(5/2)
In reality you will observe slight deviations from that because the cross sections for the photoemission are not fully identical. Therefore, if you evaluate your XPS in order to calculate a stoichiometry, check out whether your software is using a set of cross sections that includes coupling or one that doesn't.
The most popular sets are probably the Scofield and the Yeh sets; Scofield includes this effect while Yeh doesn't; on the other hand, Yeh used a slightly better calculation method, so for the choice of the set I would set the preference like this:
-you don't use peaks with coupled signals (as in most of my own XPS work) -> Yeh
-your evaluation requires the interpretation of coupled signals ->Scofield
Just saying that one is spin up and one down is wrongly overfacilitating; those coupled states are degenerate and each involves a number of energetically equivalent states (that's where the multiplicity comes from). You will only be able to differentiate between "up" and "down" cases if you apply a Zeeman or Stark effect scenario which is not the case in usual XPS.